Nextflow
Nextflow
Nextflow is a type of workflow manager, designed to be portable and reproducible. Nextflow can be used on local, HPC schedulers, AWS Batch, Google Cloud Life Sciences, and Kubernetes. Nextflow can access tool dependencies through Conda, Spack, Docker, Podman, Singularity, Modules and more.
How Do I Access Nextflow on the Cluster?
Nextflow can be accessed on the Tufts HPC Cluster with the following modules:
module load gcc nextflow/21.10.1 java/15.0.2 anaconda/2021.11
You might also have noticed that we include anaconda/2021.11 as well. This is because many nextflow pipelines you might import will use Anaconda for pipeline tool dependencies.
Running nf-core Pipelines
Speaking of importing Nextflow pipelines, it is worth mentioning nf-core pipelines. These are community developed pipelines that are available for your use! There are a number of pipelines you can use:
Danger
When Nextflow is running tasks it will create many temp files within: the ./work directory. Be very careful when running these pipelines and always be sure to clear out files in this folder to avoid running out of file storage.
We will demonstrate how to run the RNA-seq pipeline:
nextflow run nf-core/rnaseq \
--input samplesheet.csv \
--outdir <OUTDIR> \
--genome GRCh37 \
-profile docker
Where your samplesheet.csv will include:
samplesheet.csv
sample,fastq_1,fastq_2,strandedness
CONTROL_REP1,AEG588A1_S1_L002_R1_001.fastq.gz,AEG588A1_S1_L002_R2_001.fastq.gz,auto
CONTROL_REP1,AEG588A1_S1_L003_R1_001.fastq.gz,AEG588A1_S1_L003_R2_001.fastq.gz,auto
CONTROL_REP1,AEG588A1_S1_L004_R1_001.fastq.gz,AEG588A1_S1_L004_R2_001.fastq.gz,auto
However, this is a multi-faceted pipeline, with many additional parameters you can tweak for your needs. To better understand those parameters, always check the Usage docs and Parameters sections of each pipeline's documentation.